A parallel distributed approach for AI-enabled online optimization of building cooling systems with enhanced efficiency, reliability and scalability

As AI-enabled strategies are increasingly adopted for the optimization and control of building cooling systems, the efficiency and reliability of their practical implementation have become critical concerns. Conventional centralized optimization and hierarchical distributed optimization rely on central stations or coordinating agents, introducing risks of single points of failure and requiring extensive data aggregation and network transmission. This paper proposes a novel parallel distributed control strategy for AI-enabled online optimization of building cooling systems. A decentralized operational framework is developed, enabling agents spatially distributed across field stations or edge devices to independently perform optimization tasks. By leveraging dynamic task partitioning, peer-to-peer communication, and voting-based consensus mechanisms, these agents collaboratively achieve global system optimization without the need for centralized coordination. Hardware-in-the-loop tests are conducted by deploying the strategy on four real smart field control stations integrated with a virtual cooling system built using dynamic system simulation. Test results show that the proposed strategy improves scalability, operational reliability and optimization efficiency, while delivering 9.16% energy saving.